UC San Diego Graduate Student Awarded Bloomberg Data Science Ph.D. Fellowship

“Hao’s recognition as a Bloomberg Data Science Ph.D. Fellow highlights our commitment to fostering excellence and innovation at the School of Computing, Information and Data Sciences,” said SCIDS Interim Dean Rajesh Gupta, also the founding director of HDSI. “His work in advancing reasoning capabilities of LLMs demonstrates the transformative impact of interdisciplinary research in data science and artificial intelligence. We are proud to see his efforts recognized in the field.”

“Receiving the Bloomberg Data Science Fellowship is an incredible honor and a humbling recognition of my research in LLMs reasoning,” shared Hao. “This fellowship not only offers vital financial support but also provides access to a network of leading experts and valuable resources in data science. I’m truly excited and thankful for the chance to learn and contribute meaningfully to the research field through this fellowship.”

The program provides Bloomberg fellows with financial support, mentorship, career counseling, research internships and other opportunities and resources. These benefits enable recipients like Hao to focus deeply on advancing their research. Additionally, students awarded the opportunity must complete a 14-week paid summer internship at Bloomberg for each year that they are a fellow, starting in Summer 2025. Ph.D. students in the program can renew their fellowship annually for up to three years. 

For Hao, the fellowship will cover 100% of his 2024-2025 academic year tuition and award him a $42,000 stipend to cover living expenses and research expenses, such as computer hardware or attending professional conferences. With this support, Hao aims to expand the scope of his innovative research and further contribute to the field of data science by designing principled methods to enhance the reasoning ability of foundation models, especially LLMs. This work includes building a system-2 reasoning framework with a world model and a search algorithm for planning.

“My research aims to develop machine reasoning methods that emulate this human-like thought process, particularly by incorporating the concept of a world model. This could potentially lead to more advanced AI systems that make accurate, reliable decisions across fields such as healthcare, scientific discovery and beyond, boosting productivity and driving significant positive change in society,” said Hao.

When asked about how the work he’ll be leading in this fellowship will impact the UC San Diego community, Hao reflected on the broader impact of his work. 

"While my work is just one piece of the broader research happening at the university, I believe it can help foster collaboration and open up opportunities for others to explore related areas in AI," he shared. "I aim to share useful insights or tools with students and faculty, contributing to the collective efforts at UCSD."

According to the Bloomberg Data Science Ph.D. Fellowship program’s website, Bloomberg has been inviting and awarding Ph.D. students around the world working in relevant fields of study with opportunities to develop groundbreaking research publications and other forms of research since 2018.

Learn more about research and education at UC San Diego in: Artificial Intelligence